RestRserve is an R web API framework for building high-performance AND robust microservices and app backends. With Rserve backend on UNIX-like systems it is parallel by design. It will handle incoming requests in parallel - each request in a separate fork (all the credits should go to Simon Urbanek).

Quick start

Creating application is as simple as:

library(RestRserve)
app = Application$new()

app$add_get(
  path = "/health", 
  FUN = function(.req, .res) {
    .res$set_body("OK")
  })

app$add_post(
  path = "/addone", 
  FUN = function(.req, .res) {
    result = list(x = .req$body$x + 1L)
    .res$set_content_type("application/json")
    .res$set_body(result)
  })


backend = BackendRserve$new()
backend$start(app, http_port = 8080)

Test it with curl:

Autocomplete

Using convenient .req, .res names for handler arguments allows to leverage autocomplete.

Learn RestRserve

Features

  • Stable, easy to install, small number of dependencies
  • Fully featured http server with the support for URL encoded and multipart forms
  • Build safe and secure applications - RestRserve supports https, provides building blocks for basic/token authentication
  • Concise and intuitive syntax
  • Raise meaningful http errors and allows to interrupt request handling from any place of the user code
  • Well documented, comes with many examples - see inst/examples
  • Saves you from boilerplate code:
    • automatically decodes request body from the common formats
    • automatically encodes response body to the common formats
    • automatically parses URI templates (such as /get/{item_id})
    • helps to expose OpenAPI and Swagger/Redoc/Rapidoc UI
  • It is fast!

Installation

From CRAN

install.packages("RestRserve", repos = "https://cloud.r-project.org")

Docker

Debian and Alpine based images are available from docker-hub: https://hub.docker.com/r/rexyai/restrserve/

Or install specific version:

Contributing

Guidelines for filing issues / pull requests - CONTRIBUTING.md.

Acknowledgements

Known limitations

  • RestRserve is primarily tested on UNIX systems. While it works natively on Windows please don’t expect it to be as performant as on UNIX-like systems. If you really want to use it on Windows - consider to use Windows Subsystem for Linux.
  • Keep in mind that every request is handled in a separate process (fork from a parent R session). While this feature allows to handle requests in parallel it also restricts reuse of certain objects which are not fork-safe (notably database connections, rJava objects, etc)